Date of Award


Degree Type


Degree Name

Honors Thesis



First Advisor

Mike Roscoe


Objective. The purpose of this survey study was to investigate the noncausal relationships between various attribute variables of first year PA students (age, gender, ethnicity, education level, family income, and professional phase GPA) and their professionalism ratings of photos of three different outfits depicting each of five categories of attire (scrubs, casual, business casual, business, and white coat) worn in clinical settings.

Methods. An observational ecological design was utilized to measure students' perceptions of professional attire. First-year physician assistant students from Butler University in Indianapolis, Indiana were asked to rank outfits from 1= unprofessional to 7= professional using a semantic differential scale. Items were analyzed both for relationships between attribute variables and students' ratings of professional attire, and for mean intragroup differences in professionalism ratings for each dichotomized attribute variable.

Results. No significant bivariate correlations were found between any attribute variable and the mean professionalism rating for any of the five types of attire. After attribute variables were dichotomized, isolated significant intragroup mean differences were found for: age and casual clothing outfit #3 [t(45) =-2.176, p=.035]; age and white coat outfit #1 [t(32.79) =2.138, p=.040]; ethnicity and business outfit #1 [t(40) =-2.619, p= 0.12]; ethnicity and white coat outfit #3 [t(40)= - 5.234, p= .000]; GPA and business casual outfit #2 [t(45) =2.148, p=.037]; and GPA and white coat outfit #1 [t(44.135) =2.093, p=.042]. Two-variable chi square tests of independence revealed a significant relationship between trichotomized professionalism ratings for mean business casual attire and dichotomized educational level [~ (1 ,n=47) = 3.855, p=.05]. No full multiple regression models (using dichotomized attribute variables as predictors and mean professionalism ratings for each attire type as criterion variables) reached statistical significance.

Conclusions. While there were isolated mean intragroup differences and relationships that reached statistical significance, no single variable demonstrated statistical significance across all three outfits pictured for each type of attire.